EP2252088A1 - Image processing method and system - Google Patents
Image processing method and system Download PDFInfo
- Publication number
- EP2252088A1 EP2252088A1 EP09718449A EP09718449A EP2252088A1 EP 2252088 A1 EP2252088 A1 EP 2252088A1 EP 09718449 A EP09718449 A EP 09718449A EP 09718449 A EP09718449 A EP 09718449A EP 2252088 A1 EP2252088 A1 EP 2252088A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- image
- images
- feature information
- definition
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N17/00—Diagnosis, testing or measuring for television systems or their details
- H04N17/002—Diagnosis, testing or measuring for television systems or their details for television cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20224—Image subtraction
Definitions
- the present invention relates to image processing technologies, and in particular, to a method and a system for processing images.
- camera calibration is a process of obtaining geometric and optical features inside the camera, namely, internal parameters, and obtaining the three-dimensional location relation between the camera coordinate system and the space coordinate system, namely, approximation of external parameters.
- the image processing in the prior art requires the photographer to take photos from different perspectives, which brings inconvenience to the person who holds the image and to the photographer. In most circumstances, an image is obtained through cooperation of two persons. In practical experiment, using unclear images calibrating a camera leads to great errors of the calibration results or even failure of obtaining the calibration results.
- the photographer has to take multiple photos and select a clear and qualified image among them for detecting feature information, and the camera is unable to select a proper clear image automatically.
- the embodiments of the present invention provide a method and a system for processing images to select a proper and clear image automatically among video files or a series of images and automate the calibration.
- An image processing method disclosed according to a first aspect of the present invention includes:
- An image processing system disclosed according to a second aspect of the present invention includes:
- the embodiments of the present invention judge feature information and definition of a single image, and find the clear image that fulfills conditions automatically, without involving comparison between multiple images. Therefore, the images that need to be processed are reduced, image preprocessing can be performed for video files, and the subsequent video calibration is automated.
- FIG. 1 is a flowchart of an image processing method in the first embodiment of the present invention. As shown in FIG. 1 , the method includes the following blocks:
- the method in this embodiment processes obtained images, and discards one of two adjacent images if the difference between the two adjacent images is less than a preset threshold, thus ensuring distinct difference between obtained images and preventing multiple detected images from staying static.
- the method in this embodiment detects definition of every image among multiple distinctly different images, discards the unclear images, and ensures that the remaining images have distinct feature information, and are clear and mutually different. Nevertheless, in some circumstances, it is enough to obtain one image that meets the conditions. If operations related to camera calibration are performed after the image is obtained, some calibration methods require only one image, and other calibration methods require two or more images.
- the traditional camera calibration method requires at least two images. In theory, the automatic calibration of cameras requires only one image. Therefore, a certain number of clear images with feature information may be reserved as required.
- FIG. 2 is a flowchart of an image processing method in the second embodiment of the present invention. The method includes the following blocks:
- the method in this embodiment detects feature information of every image in a series of images obtained, crops the image, and judges definition of each image separately, namely, judges part with distinct feature information in an image is clear.
- the judgment about definition of an image is based on this image alone.
- image cropping is performed to maximize the percentage of the detected image feature information to the whole image. For some images, the detected feature information is little, the percentage of the feature information to the whole image is small, and the image is not easy to process. After image cropping is performed, the unnecessary peripheral zone of the obtained image is removed, and the percentage of the feature information to the image is maximized to facilitate subsequent processing.
- a series of images are obtained by photographing of the same scene, and the clearest image is found among the series of images.
- a single image is judged with respect to feature information and definition, and the clear image that conforms to conditions is found automatically, without involving comparison between multiple images. Therefore, the number of images that need to be processed is reduced, the subsequent calibration can be performed only after a certain number of qualified images are selected, and it is not necessary to process all images.
- the method provided in this embodiment can obtain clear valid images that have distinct feature information from the video files accurately, preprocess the video files, and automate the subsequent video calibration process. Moreover, because the video images or a series of frame videos can be processed automatically, one person is enough for processing the images and getting ready for subsequent calibration.
- FIG. 3 is a flowchart of an image processing method in the third embodiment of the present invention. This embodiment is similar to the embodiment shown in FIG. 2 , but differs in that: After the images are obtained in block 001 and before the feature information of the images is detected in block 002, the following steps occur:
- the detection of difference between two adjacent images may occur after the operation of detecting definition of the images (block 004 in the embodiment as shown in FIG. 2 ) or after the operation of obtaining the image (block 001 in the embodiment as shown in FIG. 3 ) or occur at the same time; and a judgment is made about whether the two images are distinctly different. If the two images are distinctly different, the two adjacent images are reserved for undergoing subsequent detection and judgment; otherwise, one of the two adjacent images is discarded. In the prior art, multiple obtained images are compared, and the clearest image is found for performing subsequent calibration calculation.
- the detection and judgment operations can be performed only for the remaining six images, and the detection stops as long as the images of the preset number meeting the requirements are detected.
- this embodiment automates the detection and treatment. Moreover, because clear images of the set number are obtained after image preprocessing, calibration calculation can be performed subsequently according to the obtained images. Because the input images of the set number have feature information and are clear, the calibration-related parameter information can be calculated effectively. Compared with the prior art, the method provided in this embodiment quickens the treatment, saves time, and enables calculation of accurate parameter information in the subsequent calibration calculation.
- FIG. 4 is a flowchart of detecting feature information in an embodiment of the present invention. As shown in FIG. 4 , the detection of feature information in this embodiment includes the following blocks:
- This embodiment uses a vertex as feature information to illustrate the operation of detecting feature information.
- feature information may be color feature information of an image, brightness feature information of image, or geometric feature information of an image. More specifically, geometric feature information may be edge or profile, vertex, straight line, circle, ellipse, or rectangle, and so on.
- the detected feature information may be a vertex, and the subsequent step 023 is to judge whether the number of actually detected vertexes is the same as the actual number of vertexes in the checkerboard. If such is the case, it indicates that the obtained image has distinct feature information; if not, the obtained image is incomplete and needs to be discarded.
- the detected feature information may be the number of circles, and the subsequent step is to judge whether the number of actually detected concentric circles is the same as the actual number of concentric circles in the template. If such is the case, it indicates that the obtained image has distinct feature information; otherwise, the obtained image is incomplete and needs to be discarded.
- different feature information may be detected for the image.
- the color feature information and the brightness feature information may be embodied as grayscale feature information, binarized image feature information, and color change feature information.
- the geometric feature information of an image may be a common straight line, circle, arc, ellipse, hyperbola, or vertex feature information (a vertex is also known as a feature point or key point).
- the brightness or chroma or gradient of the vertex changes acutely or sharply in a certain scope of the vertex.
- a peak of a black rectangle in a white background is a vertex
- a point in a black and white checkerboard is a vertex.
- a typical image is the template.
- an image may be an ordinary object other than a template, for example, an image obtained by photographing of a cup on a table.
- the detected feature information is feature information corresponding to the shape of the cup.
- the following takes a checkerboard and concentric circles as examples, which should not be construed as limitation to the present invention.
- cropping of an image detected as having feature information includes the following steps:
- FIG. 5 shows the first embodiment of image cropping.
- the image template is a checkerboard.
- the image template is a checkerboard
- four vertexes are detected through the feature information detection.
- the image may be cropped according to the detected vertex feature information, and the checkerboard is reserved to the utmost. As shown in FIG. 5 , the cropping operation includes the following steps:
- the maximum value (MaxY) and the minimum value (MinY) of the Y-coordinate of the four vertexes As shown in FIG. 5 , the maximum value is the Y-coordinate of V1 and the minimum value is the Y-coordinate of V3.
- FIG. 8 the cropped image in FIG. 8 and the uncropped image in FIG. 9 .
- the effective percentage of the checkerboard to the whole image in the uncropped image is small.
- the cropped image in FIG. 8 reserves only the checkerboard that includes all feature information, which increases the percentage of feature information to the image significantly.
- FIG. 6 shows image cropping in the second embodiment of the present invention. In this embodiment, it is assumed that the image template is concentric circles.
- the image template is concentric circles
- four concentric circles are detected through the feature information detection, as shown in FIG. 6 .
- the image may be cropped according to the detected feature information of concentric circles, and the image template is reserved to the utmost. As shown in FIG. 6 , the cropping operation includes the following steps:
- the maximum value (MaxY) and the minimum value (MinY) of the Y-coordinate of the circular boundary of the outermost concentric circle As shown in FIG. 6 , the maximum value is the Y-coordinate of V21 and the minimum value is the Y-coordinate of V23.
- the feature information may be vertexes, lines, circles, ellipses, and rectangles, and the corresponding boundary and image cropping are of a little difference.
- FIG. 7 is a flowchart of an image processing method in the four embodiment of the present invention. This embodiment gives a detailed flowchart. As shown in FIG. 7 , still taking the checkerboard image template as an example, the method provided in this embodiment includes the following blocks:
- This block further includes:
- the obtained image may also be an image detected in real time, as obtained in block 01.
- Block 01 Capture by the camera a series of images in real time when the image with a template is shaking.
- This block further includes:
- a video file is obtained by photographing, and is stored in a computer hard disk in a format similar to AVI.
- the video file includes multiple images.
- continuous frames (the same image) are obtained through real-time photographing, and are stored in the computer cache.
- the image may be obtained by other means. For example, one person moves while operating the camera, and the checkerboard is stationary; or an image is obtained by photographing directly without checkerboard.
- the man-caused factors may exist in the photographing, for example, the camera is shaking, or the camera incurs unintentional interference. Consequently, the image is not clear, and is not suitable for calibration.
- subsequent blocks are performed, including:
- FIG. 10 is a grayscale image of a clear image
- FIG. 10 is a grayscale image corresponding to a vague image.
- the lines in the grayscale image corresponding to the clear image are thin, and the lines in FIG. 11 are thick.
- the preset threshold there are relatively many white pixel points in the image, and the lines are thick, as shown in FIG. 11 .
- the corresponding image is vague, and needs to be excluded for subsequent detection.
- the block of judging definition of the image is equivalent to block 042' in FIG. 7 .
- the preset definition criterion is the percentage of the feature information (such as edge or profile information) corresponding to the clearest image template to the cropped image. If the actually obtained cropped image is out of the set definition criterion, the image is vague. Alternatively, the total number of white pixel points of feature information in the cropped image is compared with the set definition criterion. If the total sum of white pixel points exceeds the set definition criterion, the image is vague, and the corresponding lines are thick, as shown in FIG. 11 . After the image is determined as clear, the following blocks occur:
- the detection of difference between two adjacent images may occur after the operation of judging definition of the images or after the operation of obtaining the image or occur at the same time.
- a proper threshold is set, and a judgment is made about whether the two adjacent images are distinctly different; if the two images are distinctly different, the two adjacent images are reserved for undergoing subsequent detection and judgment; otherwise, one of the two adjacent images is discarded.
- the difference between two adjacent images is detected in order to prevent the detected images from staying static and unchanged.
- the operation of judging whether the difference between two adjacent images is greater than the preset threshold includes these steps:
- Block 06 Obtain a sequence of clear calibration images, namely, multiple images meeting conditions (feature information exists, and the images are clear and cropped).
- Block 07 Judge whether the number of images in the image sequence is greater than 3; if so, go to block 08; otherwise, go to block 01 or 01' to obtain more images. If a two-dimensional or three-dimensional template is used to calibrate the camera, judge whether the final number of the effective calibration images conforms to the required number of calibration images. In this embodiment, the required number of calibration images is set to 3.
- Block 08 Calibrate the video obtaining device according to the three clear images obtained and the corresponding feature information.
- Block 09 Obtain internal and external parameters of the video obtaining device.
- the image definition judgment in this embodiment is based on the Fourier optical theory:
- the extent of definition or focusing of an image primarily depends on the number of high frequency components in the light intensity distribution. Small number of high frequency components lead to vague images and large number of high frequency components lead to clear images.
- This embodiment uses the content of high frequency components in the light intensity distribution of an image as a main basis of assessing the definition or focusing of the image.
- an image has an edge part, when the camera is in full focus, the image is clear, and the most high frequency components that include the edge information exist; when the camera is out of focus, the image is vague, and less high frequency components exist. Therefore, the number of high frequency components of edge information of the image can be used as a basis for judging whether the image is clear or in focus.
- the definition judgment in this embodiment is different from the method of judging image definition or automatic focusing in the prior art.
- a series of images are obtained by photographing of the same scene, and the clearest image is found among them for adjusting the focal length of the camera automatically. That is, in the prior art, comparison is made between a series of images, and the clearest image is selected among them as a basis for judging whether the focal length is adjusted to the ideal location.
- a judgment is made about whether an image alone is clear, or whether the part with distinct feature information in the image is clear, regardless of whether other parts of the image are clear.
- the image definition judgment in this embodiment is not based on a series of images, but is based on a single image alone.
- FIG. 12 shows an image processing system in the first embodiment of the present invention. As shown in FIG. 12 , the system in this embodiment includes:
- FIG. 13 shows an image processing system in the second embodiment of the present invention. As shown in FIG. 13 , the system in this embodiment includes:
- the system in this embodiment may further include: a calibrating module 7, connected with the image number judging module 6, and adapted to: obtain internal and external parameters of the video obtaining device according to the set number of obtained clear images and the corresponding feature information, and calibrate the video obtaining device.
- a calibrating module 7 connected with the image number judging module 6, and adapted to: obtain internal and external parameters of the video obtaining device according to the set number of obtained clear images and the corresponding feature information, and calibrate the video obtaining device.
- FIG. 14 shows an image processing system in the third embodiment of the present invention. As shown in FIG. 14 , the system in this embodiment includes:
- FIG. 15 shows an image processing system in the fourth embodiment of the present invention. As shown in FIG. 15 , the system in this embodiment includes:
- the system in this embodiment may further include: a calibrating module 7, connected with the image number judging module 6, and adapted to: obtain internal and external parameters of the video obtaining device according to the set number of obtained clear images and the corresponding feature information, and calibrate the video obtaining device.
- a calibrating module 7 connected with the image number judging module 6, and adapted to: obtain internal and external parameters of the video obtaining device according to the set number of obtained clear images and the corresponding feature information, and calibrate the video obtaining device.
- FIG. 12 to FIG. 15 correspond to the method embodiments in FIG. 1 to FIG. 7 , and provide the functions and effects similar to the embodiments in FIG. 1 to FIG. 7 , and implement image preprocessing to facilitate subsequent calibration operation.
- the present invention has different embodiments.
- the technical solution under the present invention is described above with reference to FIG. 1 to FIG. 15 , but the present invention is not limited to the method processes or system structures described above. Persons of ordinary skill in the art understand that the implementation solutions described above are only preferred and exemplary solutions.
- the program may be stored in a computer readable storage medium. When the program runs, the steps of the method according to the embodiments of the present invention are performed.
- the storage medium may be a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or a Compact Disk Read-Only Memory (CD-ROM).
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
Description
- This application claims priority to Chinese Patent Application No.
200810082736.2 - The present invention relates to image processing technologies, and in particular, to a method and a system for processing images.
- In a computer visual system, camera calibration is a process of obtaining geometric and optical features inside the camera, namely, internal parameters, and obtaining the three-dimensional location relation between the camera coordinate system and the space coordinate system, namely, approximation of external parameters.
- No matter what camera calibration method is adopted, obtaining clear images automatically is a foundation of successful calibration. If the images are not clear, the calibration result is deteriorated.
- In the process of developing the present invention, the inventor finds that the prior art does not define how to obtain clear images automatically. The image processing in the prior art requires the photographer to take photos from different perspectives, which brings inconvenience to the person who holds the image and to the photographer. In most circumstances, an image is obtained through cooperation of two persons. In practical experiment, using unclear images calibrating a camera leads to great errors of the calibration results or even failure of obtaining the calibration results. In the prior art, the photographer has to take multiple photos and select a clear and qualified image among them for detecting feature information, and the camera is unable to select a proper clear image automatically.
- The embodiments of the present invention provide a method and a system for processing images to select a proper and clear image automatically among video files or a series of images and automate the calibration.
- An image processing method disclosed according to a first aspect of the present invention includes:
- obtaining video images or capturing a series of images in real time;
- detecting difference between two adjacent images, and reserving one of the images when the difference is less than a preset threshold;
- detecting feature information of the obtained image; and
- performing definition-related treatment for the image detected as having the feature information, comparing definition of the image with a preset definition criterion, and reserving clear image.
- An image processing system disclosed according to a second aspect of the present invention includes:
- an image obtaining module, adapted to obtain video images or capture a series of images in real time;
- a feature information detecting module, connected with the image obtaining module, and adapted to detect feature information of the obtained video images or the series of images and obtain a image that has distinct feature information;
- an image definition judging module, connected with the feature information detecting module and the image obtaining module, and adapted to: perform definition-related treatment for the images that have the feature information, compare definition of each image with a preset definition criterion, and reserve clear images; and
- an image difference detecting module, connected with the image definition judging module, and adapted to store a current clear image if difference between a current image and a previous stored image is greater than a preset threshold.
- Another image processing system disclosed in an embodiment of the present invention includes:
- an image obtaining module, adapted to obtain video images or capture a series of images in real time;
- an image difference detecting module, connected with the image obtaining module, and adapted to store two adjacent images when difference between the two adjacent images is greater than a preset threshold;
- a feature information detecting module, connected with the image obtaining module, and adapted to detect feature information of the obtained video images or the series of images and obtain images that have distinct feature information; and
- an image definition judging module, connected with the feature information detecting module and the image obtaining module, and adapted to: perform definition-related treatment for the images that have the feature information, compare definition of each image with a preset definition criterion; when a current image is unclear, discard the current image so that the image obtaining module continues to select a next image.
- The embodiments of the present invention judge feature information and definition of a single image, and find the clear image that fulfills conditions automatically, without involving comparison between multiple images. Therefore, the images that need to be processed are reduced, image preprocessing can be performed for video files, and the subsequent video calibration is automated.
-
-
FIG. 1 is a flowchart of an image processing method in the first embodiment of the present invention; -
FIG. 2 is a flowchart of an image processing method in the second embodiment of the present invention; -
FIG. 3 is a flowchart of an image processing method in the third embodiment of the present invention; -
FIG. 4 is a flowchart of detecting feature information in an embodiment of the present invention; -
FIG. 5 shows image cropping in the first embodiment of the present invention; -
FIG. 6 shows image cropping in the second embodiment of the present invention; -
FIG. 7 is a flowchart of an image processing method in the fourth embodiment of the present invention; -
FIG. 8 shows a cropped image in an image processing method in an embodiment of the present invention; -
FIG. 9 shows an uncropped image in an image processing method in an embodiment of the present invention; -
FIG. 10 shows a gradient map of a cropped image in an image processing method in an embodiment of the present invention; -
FIG. 11 shows a gradient map of an uncropped vague image in an image processing method in an embodiment of the present invention; -
FIG. 12 shows an image processing system in the first embodiment of the present invention; -
FIG. 13 shows an image processing system in the second embodiment of the present invention; -
FIG. 14 shows an image processing system in the third embodiment of the present invention;
and -
FIG. 15 shows an image processing system in the fourth embodiment of the present invention. - The technical solution under the present invention is elaborated below with reference to accompanying drawings and preferred embodiments.
-
FIG. 1 is a flowchart of an image processing method in the first embodiment of the present invention. As shown inFIG. 1 , the method includes the following blocks: - Block 0001: Obtain video images or capture a series of images in real time.
- Block 0002: Detect difference between two adjacent images.
- Block 0003: Judge whether the difference between the two images is less than a preset threshold; if so, go to block 0004; if not, go to block 0005.
- Block 0004: Discard one of the images, and reserve the other.
- Block 0005: Detect feature information of the obtained image.
- Block 0006: Perform definition detection for the images detected as having the feature information, compare definition of the images with a preset definition criterion, and reserve clear images.
- The method in this embodiment processes obtained images, and discards one of two adjacent images if the difference between the two adjacent images is less than a preset threshold, thus ensuring distinct difference between obtained images and preventing multiple detected images from staying static. The method in this embodiment detects definition of every image among multiple distinctly different images, discards the unclear images, and ensures that the remaining images have distinct feature information, and are clear and mutually different. Nevertheless, in some circumstances, it is enough to obtain one image that meets the conditions. If operations related to camera calibration are performed after the image is obtained, some calibration methods require only one image, and other calibration methods require two or more images. The traditional camera calibration method requires at least two images. In theory, the automatic calibration of cameras requires only one image. Therefore, a certain number of clear images with feature information may be reserved as required.
-
FIG. 2 is a flowchart of an image processing method in the second embodiment of the present invention. The method includes the following blocks: - Block 001: Obtain video images or capture a series of images in real time.
- Block 002: Detect feature information of an image among the video images or the series of frame images.
- Block 003: Crop the image that are detected as having feature information.
- Block 004: Perform definition detection for the cropped image (perform definition-related treatment, and compare definition of the image with a preset definition criterion), and discard the unclear image.
- Block 005: Judge whether difference between the current image and the previous image is greater than a preset threshold; if so, it means that the two images are distinctly different, and go to block 006; otherwise, the two images are scarcely different and go to block 051 to prevent the detected scene from staying static.
- Block 051: Because the two images are scarcely different, discard one of the images, reserve only one image for participating in the subsequent treatment, and go on to block 002 to read the next image.
- Block 006: Store the current image.
- Block 007: Judge whether the number of images reaches a preset value. For example, set the value to 3 because at least three different images are required for subsequent internal parameter operation in camera calibration; or set the value to others such as 4 in order to reserve a redundancy. Go to block 008 when the number of images reaches a preset value; or go to block 002 to read the next image if the number of images that meet the conditions is less than the preset value.
- Block 008: Obtain clear images of the number according to the set value.
- The method in this embodiment detects feature information of every image in a series of images obtained, crops the image, and judges definition of each image separately, namely, judges part with distinct feature information in an image is clear. The judgment about definition of an image is based on this image alone. In this embodiment, image cropping is performed to maximize the percentage of the detected image feature information to the whole image. For some images, the detected feature information is little, the percentage of the feature information to the whole image is small, and the image is not easy to process. After image cropping is performed, the unnecessary peripheral zone of the obtained image is removed, and the percentage of the feature information to the image is maximized to facilitate subsequent processing.
- In the prior art, a series of images are obtained by photographing of the same scene, and the clearest image is found among the series of images. In the foregoing embodiment of the present invention, a single image is judged with respect to feature information and definition, and the clear image that conforms to conditions is found automatically, without involving comparison between multiple images. Therefore, the number of images that need to be processed is reduced, the subsequent calibration can be performed only after a certain number of qualified images are selected, and it is not necessary to process all images. The method provided in this embodiment can obtain clear valid images that have distinct feature information from the video files accurately, preprocess the video files, and automate the subsequent video calibration process. Moreover, because the video images or a series of frame videos can be processed automatically, one person is enough for processing the images and getting ready for subsequent calibration.
-
FIG. 3 is a flowchart of an image processing method in the third embodiment of the present invention. This embodiment is similar to the embodiment shown inFIG. 2 , but differs in that: After the images are obtained inblock 001 and before the feature information of the images is detected inblock 002, the following steps occur: - Block 010: Judge whether difference between two adjacent images is greater than a preset threshold; if so, it means that the two images are distinctly different, and go to block 011; otherwise, the two images are scarcely different, and go to block 012 to prevent the detected scene from staying static.
- Block 011: Store the two images, and go to block 002.
- Block 012: Because the two images are scarcely different, discard one of the images, reserve only one image for participating in the subsequent treatment, and go to block 002.
- In the embodiments elaborated in
FIG. 1 to FIG. 3 , the detection of difference between two adjacent images may occur after the operation of detecting definition of the images (block 004 in the embodiment as shown inFIG. 2 ) or after the operation of obtaining the image (block 001 in the embodiment as shown inFIG. 3 ) or occur at the same time; and a judgment is made about whether the two images are distinctly different. If the two images are distinctly different, the two adjacent images are reserved for undergoing subsequent detection and judgment; otherwise, one of the two adjacent images is discarded. In the prior art, multiple obtained images are compared, and the clearest image is found for performing subsequent calibration calculation. In the embodiments above, if two images obtained are not distinctly different, one of the images is discarded to prevent several images from staying static, and ensure distinct difference between the final images for calibration calculation and that definition requirements are met. For example, if there are ten images obtained by photographing or ten frames of a video image, but four of them are scarcely different, the detection and judgment operations can be performed only for the remaining six images, and the detection stops as long as the images of the preset number meeting the requirements are detected. - In this embodiment, several images that are unclear are discarded before calibration calculation, and the obtained images of a set number are used to perform subsequent calibration. In the example given above, if the first three of the six detected images conform to requirements (such as feature information and definition), the last three images do not need to be processed; if two of the first three images in the six detected images conform to requirements (such as feature information and definition), it is necessary to detect the last three images, and the images of the preset number are finally obtained. In this way, three to six times of treatment are enough for completing the preprocessing. Compared with the prior art, this embodiment automates the detection and treatment. Moreover, because clear images of the set number are obtained after image preprocessing, calibration calculation can be performed subsequently according to the obtained images. Because the input images of the set number have feature information and are clear, the calibration-related parameter information can be calculated effectively. Compared with the prior art, the method provided in this embodiment quickens the treatment, saves time, and enables calculation of accurate parameter information in the subsequent calibration calculation.
-
FIG. 4 is a flowchart of detecting feature information in an embodiment of the present invention. As shown inFIG. 4 , the detection of feature information in this embodiment includes the following blocks: - Block 021: Read every frame sequentially or read an image captured in real time.
- Block 022: Detect vertex information in the image.
- Block 023: Judge whether the number of actually detected vertexes is the same as the theoretic number of vertexes; if so, go on with the subsequent blocks; if not, discard the currently detected image, and go to block 021 to read the image next to the currently detected image.
- This embodiment uses a vertex as feature information to illustrate the operation of detecting feature information. However, those skilled in the art understand that feature information may be color feature information of an image, brightness feature information of image, or geometric feature information of an image. More specifically, geometric feature information may be edge or profile, vertex, straight line, circle, ellipse, or rectangle, and so on. If the template of the image is a checkerboard, the detected feature information may be a vertex, and the
subsequent step 023 is to judge whether the number of actually detected vertexes is the same as the actual number of vertexes in the checkerboard. If such is the case, it indicates that the obtained image has distinct feature information; if not, the obtained image is incomplete and needs to be discarded. If the template of the image is concentric circles, the detected feature information may be the number of circles, and the subsequent step is to judge whether the number of actually detected concentric circles is the same as the actual number of concentric circles in the template. If such is the case, it indicates that the obtained image has distinct feature information; otherwise, the obtained image is incomplete and needs to be discarded. In practice, different feature information may be detected for the image. The color feature information and the brightness feature information may be embodied as grayscale feature information, binarized image feature information, and color change feature information. The geometric feature information of an image may be a common straight line, circle, arc, ellipse, hyperbola, or vertex feature information (a vertex is also known as a feature point or key point). As regards the meaning of the vertex, the brightness or chroma or gradient of the vertex changes acutely or sharply in a certain scope of the vertex. For example, a peak of a black rectangle in a white background is a vertex, and a point in a black and white checkerboard is a vertex. Nevertheless, a typical image is the template. Those skilled in the art understand that an image may be an ordinary object other than a template, for example, an image obtained by photographing of a cup on a table. In this case, the detected feature information is feature information corresponding to the shape of the cup. For ease of illustration, the following takes a checkerboard and concentric circles as examples, which should not be construed as limitation to the present invention. - In the foregoing embodiment, cropping of an image detected as having feature information includes the following steps:
- obtain coordinates of feature information of the image according to the obtained feature information;
- obtain the maximum value and the minimum value of the X-coordinate and the Y-coordinate of each point corresponding to the feature information, and obtain the image suitable for cropping;
- obtain the boundary formed by the points corresponding to the feature information according to the coordinates of the points corresponding to the feature information of the image;
- cut away the area outside the boundary formed by the points and inside the image suitable for cropping; and
- obtain a cropped image according to only the image information inside the boundary formed by the points.
- The image cropping operation is shown in
FIG. 5 andFIG. 6 .FIG. 5 shows the first embodiment of image cropping. In this embodiment, it is assumed that the image template is a checkerboard. - If the image template is a checkerboard, four vertexes are detected through the feature information detection. The image may be cropped according to the detected vertex feature information, and the checkerboard is reserved to the utmost. As shown in
FIG. 5 , the cropping operation includes the following steps: - (1) Obtain coordinates of four vertexes V1-V4 of the detected checkerboard.
- (2) Obtain the maximum value (MaxX) and the minimum value (MinX) of the X-coordinate of the four vertexes. As shown in
FIG. 5 , the maximum value is the X-coordinate of V2 and the minimum value is the X-coordinate of V4. - Obtain the maximum value (MaxY) and the minimum value (MinY) of the Y-coordinate of the four vertexes. As shown in
FIG. 5 , the maximum value is the Y-coordinate of V1 and the minimum value is the Y-coordinate of V3. - (3) Obtain the image suitable for cropping according to step (2). For example, the outermost rectangle in
FIG. 5 is the image suitable for cropping. - (4) According to the four vertexes V1, V2, V3, and V4, create a coordinate equation of four edges L1, L2, L3, and L4 which make up a rectangle V1V2V3V4.
- (5) According to the equation of the four edges, set black for all image points inside the image suitable for cropping and outside the boundary of the rectangle formed by the four edges.
- (6) Obtain a final cropped image which reserves only the internal image information inside the boundary formed by the four edges.
- More specifically, see the cropped image in
FIG. 8 and the uncropped image inFIG. 9 . As revealed byFIG. 9 , the effective percentage of the checkerboard to the whole image in the uncropped image is small. The cropped image inFIG. 8 reserves only the checkerboard that includes all feature information, which increases the percentage of feature information to the image significantly. -
FIG. 6 shows image cropping in the second embodiment of the present invention. In this embodiment, it is assumed that the image template is concentric circles. - If the image template is concentric circles, four concentric circles are detected through the feature information detection, as shown in
FIG. 6 . The image may be cropped according to the detected feature information of concentric circles, and the image template is reserved to the utmost. As shown inFIG. 6 , the cropping operation includes the following steps: - (1) Obtain coordinates of the circular boundary of the outmost concentric center among the four detected concentric circles.
- (2) Obtain the maximum value (MaxX) and the minimum value (MinX) of the X-coordinate of the circular boundary of the outermost concentric circle. As shown in
FIG. 6 , the maximum value is the X-coordinate of V24 and the minimum value is the X-coordinate of V22. - Obtain the maximum value (MaxY) and the minimum value (MinY) of the Y-coordinate of the circular boundary of the outermost concentric circle. As shown in
FIG. 6 , the maximum value is the Y-coordinate of V21 and the minimum value is the Y-coordinate of V23. - (3) Obtain the image suitable for cropping according to step (2). For example, the rectangle in
FIG. 5 is the image suitable for cropping. - (4) According to the four vertexes V1, V2, V3, and V4, create four arcs L21, L22, L23, and L24 which make up a circle.
- (5) Obtain the boundary of the circle according to step (4), set all image points inside the images suitable for cropping in step (2) and outside the boundary of the circle to be black.
- (6) Obtain a final cropped image which reserves only the image information of the four concentric circles.
- Those skilled in the art understand that for different feature information, the corresponding image cropping operation differs. The feature information may be vertexes, lines, circles, ellipses, and rectangles, and the corresponding boundary and image cropping are of a little difference.
-
FIG. 7 is a flowchart of an image processing method in the four embodiment of the present invention. This embodiment gives a detailed flowchart. As shown inFIG. 7 , still taking the checkerboard image template as an example, the method provided in this embodiment includes the following blocks: - Block 01': Obtain a video file which includes a checkerboard.
- This block further includes:
- powering on the camera to be ready for photographing;
- letting one person hold a checkerboard and shake the checkerboard template in front of the camera; and
- obtaining a video file which includes the checkerboard.
- The obtained image may also be an image detected in real time, as obtained in
block 01. - Block 01: Capture by the camera a series of images in real time when the image with a template is shaking.
- This block further includes:
- powering on the camera to be ready for photographing;
- letting one person stand in front of the camera and shake a checkerboard;
- capturing the images in real time by using the camera, and detecting and processing the images; and
- obtaining a series of images.
- Through block 01', a video file is obtained by photographing, and is stored in a computer hard disk in a format similar to AVI. The video file includes multiple images. Through
block 01, continuous frames (the same image) are obtained through real-time photographing, and are stored in the computer cache. Nevertheless, the image may be obtained by other means. For example, one person moves while operating the camera, and the checkerboard is stationary; or an image is obtained by photographing directly without checkerboard. However, the man-caused factors may exist in the photographing, for example, the camera is shaking, or the camera incurs unintentional interference. Consequently, the image is not clear, and is not suitable for calibration. - After one or more frames/pages of images are obtained, subsequent blocks are performed, including:
- Block 021: Read every frame or read an image captured in real time.
- Block 022: Check for vertex feature information.
- Block 023: Judge whether the number of detected vertexes is equal to the actual number of vertexes; if so, go to block 03; otherwise, go to block 021.
- Block 03: Crop the image, namely, block 03' shown in
FIG. 7 : crop the image according to the checkerboard in the position of the detected four vertexes, and obtain the images that have distinct feature information, where the number of the images is consistent with the feature information of the template. For details, see the description about the embodiments shown inFIG. 5 andFIG. 6 . - Block 041: Calculate the gradient of the cropped image.
- Block 042: Calculate and check whether the image is clear. If so, go to block 05 to detect difference; otherwise, go to block 021 to read the next image.
- Block 041 and block 042 are blocks of judging definition of the image, and are to judge whether the image is clear or in focus according to content of the high frequency components of edge information in the feature information of the image. The operation of detecting definition of the image is detailed below:
- A. Obtain a cropped image according to block 03 in this embodiment. If the image is clear, gray the cropped image (shown in
FIG. 8 ), and calculate the gradient value (for example, perform Laplace transformation for the image) of each point of the cropped image. As a result, output a grayscale image which includes only the edge information, as shown inFIG. 10 . - B. Calculate the absolute value of the gradient value of the image. Sometimes the gradient is not positive, and the value of each pixel point of the grayscale image ranges from 0 to 255. Therefore, the absolute value of the gradient value needs to be calculated out, and converted into a grayscale image.
- C. Make statistics to obtain the total sum of pixels of white lines of edge information in the grayscale image corresponding to the clear image, or divide the pixel sum by the size of the cropped image (namely, total number of pixels of the image), and set a threshold T1 as a definition criterion, where T1 is the total sum of white pixel points or the percentage of the total sum of white pixel points to the whole image.
- D. Enlarge the threshold. As shown in
FIG. 10 , noise points exist in each white grid.
Therefore, denoising may be performed, or the threshold T1 needs to be multiplied properly. For different scenes, the enlarging multiple differs. If the denoising effect is very good, no enlarging is required. If the denoising effect is poor and there are many noise points, the enlarging multiple is about two-fold. - E. Binarize the cropped image, and obtain the total sum of white pixel points of all edge information in the binarized image or the percentage of the edge information to the whole image.
- F. Compare the definition threshold T1 with the total sum of white pixel points of all edge information in the binarized image or the percentage of the edge information to the whole image. If the latter is greater than T1, the image is vague, and needs to be discarded; otherwise, the image is clear and needs to be reserved.
- A. Obtain a cropped image according to block 03 in this embodiment. If the image is clear, gray the cropped image (shown in
- The edge information (namely, display of the gradient map) in the grayscale image is shown in
FIG. 10 andFIG. 11 .FIG. 10 is a grayscale image of a clear image, andFIG. 10 is a grayscale image corresponding to a vague image. Evidently, the lines in the grayscale image corresponding to the clear image are thin, and the lines inFIG. 11 are thick. If the total sum of white pixel points calculated out in step F is greater than the preset threshold, there are relatively many white pixel points in the image, and the lines are thick, as shown inFIG. 11 . In this case, the corresponding image is vague, and needs to be excluded for subsequent detection. - The block of judging definition of the image is equivalent to block 042' in
FIG. 7 . The preset definition criterion is the percentage of the feature information (such as edge or profile information) corresponding to the clearest image template to the cropped image. If the actually obtained cropped image is out of the set definition criterion, the image is vague. Alternatively, the total number of white pixel points of feature information in the cropped image is compared with the set definition criterion. If the total sum of white pixel points exceeds the set definition criterion, the image is vague, and the corresponding lines are thick, as shown inFIG. 11 . After the image is determined as clear, the following blocks occur: - Block 05: Judge whether the difference between two adjacent images is greater than a preset threshold; if so, go to block 06; otherwise, go to block 021.
- As described in the embodiments in
FIG. 1 to FIG. 3 , the detection of difference between two adjacent images may occur after the operation of judging definition of the images or after the operation of obtaining the image or occur at the same time. In the difference detection, a proper threshold is set, and a judgment is made about whether the two adjacent images are distinctly different; if the two images are distinctly different, the two adjacent images are reserved for undergoing subsequent detection and judgment; otherwise, one of the two adjacent images is discarded. The difference between two adjacent images is detected in order to prevent the detected images from staying static and unchanged. In this embodiment, the operation of judging whether the difference between two adjacent images is greater than the preset threshold includes these steps: - 1. Obtain information about vertexes of two adjacent images.
- 2. Calculate the difference between the coordinate of the vertex in the corresponding position of the previous image and that of the next image.
- 3. Calculate the sum of the coordinate differences of all corresponding vertexes, and calculate the mean value of them. Set a threshold of this mean value. If the difference exceeds the threshold, it indicates that the two adjacent images are distinctly different and both need to be reserved; otherwise, one of the images needs to be discarded.
- Block 06: Obtain a sequence of clear calibration images, namely, multiple images meeting conditions (feature information exists, and the images are clear and cropped).
- Block 07: Judge whether the number of images in the image sequence is greater than 3; if so, go to block 08; otherwise, go to block 01 or 01' to obtain more images. If a two-dimensional or three-dimensional template is used to calibrate the camera, judge whether the final number of the effective calibration images conforms to the required number of calibration images. In this embodiment, the required number of calibration images is set to 3.
- Block 08: Calibrate the video obtaining device according to the three clear images obtained and the corresponding feature information.
- Block 09: Obtain internal and external parameters of the video obtaining device.
- The image definition judgment in this embodiment is based on the Fourier optical theory: The extent of definition or focusing of an image primarily depends on the number of high frequency components in the light intensity distribution. Small number of high frequency components lead to vague images and large number of high frequency components lead to clear images. This embodiment uses the content of high frequency components in the light intensity distribution of an image as a main basis of assessing the definition or focusing of the image.
- Because an image has an edge part, when the camera is in full focus, the image is clear, and the most high frequency components that include the edge information exist; when the camera is out of focus, the image is vague, and less high frequency components exist. Therefore, the number of high frequency components of edge information of the image can be used as a basis for judging whether the image is clear or in focus.
- The definition judgment in this embodiment is different from the method of judging image definition or automatic focusing in the prior art. In the method of judging image definition or automatic focusing in the prior art, a series of images are obtained by photographing of the same scene, and the clearest image is found among them for adjusting the focal length of the camera automatically. That is, in the prior art, comparison is made between a series of images, and the clearest image is selected among them as a basis for judging whether the focal length is adjusted to the ideal location. In this embodiment, in the image processing before video calibration, a judgment is made about whether an image alone is clear, or whether the part with distinct feature information in the image is clear, regardless of whether other parts of the image are clear. Meanwhile, the image definition judgment in this embodiment is not based on a series of images, but is based on a single image alone.
-
FIG. 12 shows an image processing system in the first embodiment of the present invention. As shown inFIG. 12 , the system in this embodiment includes: - an image obtaining module 1, adapted to obtain video images or capture a series of images in real time;
- a feature
information detecting module 2, connected with the image obtaining module 1, and adapted to detect feature information of the obtained video images or the series of images and obtain images that have distinct feature information; - an image
definition judging module 40, connected with the featureinformation detecting module 2 and the image obtaining module 1, and adapted to: perform definition-related treatment for the images that have the feature information, compare definition of each image with a preset definition criterion, and reserve clear images; and - an image
difference detecting module 5, connected with the imagedefinition judging module 40, and adapted to store a current clear image if difference between a current image and a previous stored image is greater than a preset threshold. -
FIG. 13 shows an image processing system in the second embodiment of the present invention. As shown inFIG. 13 , the system in this embodiment includes: - an image obtaining module 1, adapted to obtain video images or capture a series of images in real time;
- a feature
information detecting module 2, connected with the image obtaining module 1, and adapted to detect feature information of the obtained video images or the series of images and obtain images that have distinct feature information; - an
image cropping module 3, connected with the featureinformation detecting module 2, and adapted to crop the images that have the feature information; - an image
definition judging module 4, connected with theimage cropping module 3 and the image obtaining module 1, and adapted to: perform definition-related treatment for the cropped images, compare definition of the image with a preset definition criterion, and reserve the clear image; - an image
difference detecting module 5, connected with the imagedefinition judging module 4, and adapted to: judge whether difference between a current image and a previous stored image is less than a preset threshold; if so, reserve one of the images; otherwise, store the current image; and - an image
number judging module 6, connected with the imagedifference detecting module 5 and the image obtaining module 1, and adapted to: detect the number of stored images, and make the image obtaining module select the next image for subsequent operations if the number of stored images does not reach the set number. - The system in this embodiment may further include: a calibrating
module 7, connected with the imagenumber judging module 6, and adapted to: obtain internal and external parameters of the video obtaining device according to the set number of obtained clear images and the corresponding feature information, and calibrate the video obtaining device. -
FIG. 14 shows an image processing system in the third embodiment of the present invention. As shown inFIG. 14 , the system in this embodiment includes: - an image obtaining module 1, adapted to obtain video images or capture a series of images in real time;
- an image
difference detecting module 5, connected with the image obtaining module 1, and adapted to store two adjacent images when difference between the two adjacent images is greater than a preset threshold; - a feature
information detecting module 2, connected with the image obtaining module 1, and adapted to detect feature information of the obtained video images or the series of images and obtain images that have distinct feature information; and - an image
definition judging module 4, connected with the featureinformation detecting module 2 and the image obtaining module 1, and adapted to: perform definition-related treatment for the images that have the feature information, compare definition of each image with a preset definition criterion; when a current image is unclear, discard the current image so that the image obtaining module continues to select a next image. -
FIG. 15 shows an image processing system in the fourth embodiment of the present invention. As shown inFIG. 15 , the system in this embodiment includes: - an image obtaining module 1, adapted to obtain video images or capture a series of images in real time;
- an image
difference detecting module 5, connected with the image obtaining module 1, and adapted to store two adjacent images when difference between the two adjacent images is greater than a preset threshold; - a feature
information detecting module 2, connected with the image obtaining module 1, and adapted to detect feature information of the obtained video images or the series of images and obtain images that have distinct feature information; - an
image cropping module 3, connected with the featureinformation detecting module 2, and adapted to crop the images that have the feature information; - an image
definition judging module 4, connected with theimage cropping module 3 and the image obtaining module 1, and adapted to: perform definition-related treatment for the cropped images, compare definition of the image with a preset definition criterion, and reserve the clear image; and - an image
number judging module 6, connected with the image definition module and the image obtaining module 1, and adapted to: detect the number of stored images, and make the image obtaining module select the next image for subsequent operations if the number of stored images does not reach the set number. - The system in this embodiment may further include: a calibrating
module 7, connected with the imagenumber judging module 6, and adapted to: obtain internal and external parameters of the video obtaining device according to the set number of obtained clear images and the corresponding feature information, and calibrate the video obtaining device. - The embodiments in
FIG. 12 to FIG. 15 correspond to the method embodiments inFIG. 1 to FIG. 7 , and provide the functions and effects similar to the embodiments inFIG. 1 to FIG. 7 , and implement image preprocessing to facilitate subsequent calibration operation. - The present invention has different embodiments. The technical solution under the present invention is described above with reference to
FIG. 1 to FIG. 15 , but the present invention is not limited to the method processes or system structures described above. Persons of ordinary skill in the art understand that the implementation solutions described above are only preferred and exemplary solutions. - Person having ordinary skill in the art may understand that all or part of the steps of the method according to the embodiments of the present invention may be implemented by a program instructing relevant hardware. The program may be stored in a computer readable storage medium. When the program runs, the steps of the method according to the embodiments of the present invention are performed. The storage medium may be a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or a Compact Disk Read-Only Memory (CD-ROM).
- It should be noted that the above embodiments are merely provided for elaborating the technical solutions of the present invention, but not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, it is apparent that person having ordinary skill in the art may make modifications to the technical solutions described in the above embodiments, or may make some equivalent replacements to some of the technical features without making the nature of the corresponding technical solutions depart from the spirit and scope of various embodiments of the invention.
Claims (21)
- An image processing method, comprising:obtaining video images or capturing a series of images in real time;detecting difference between two adjacent images, and reserving one of the images when the difference is less than a preset threshold;detecting feature information of the obtained image; andperforming definition-related treatment for the image detected as having the feature information, comparing definition of the image with a preset definition criterion, and reserving clear image.
- The image processing method according to claim 1, wherein: before performing definition-related treatment for the images detected as having the feature information, the method further comprises:cropping the images detected as having the feature information so as to increase percentage of the detected feature information to the whole image.
- The image processing method according to claim 1, wherein: after reserving the clear image, the method further comprises:continuing to obtain the video images or capture a series of the images in real time if the number of the reserved images does not reach a set number.
- The image processing method according to any one of claims 1-3, further comprising:performing camera calibration according to the obtained clear image and corresponding feature information, and obtaining internal and external parameters of a camera.
- The image processing method according to claim 1, wherein the detecting of the feature information of the obtained images comprises:detecting color feature information, brightness feature information, and/or geometric feature information of the image, wherein: the color feature information and the brightness feature information comprise grayscale feature information, binarized image feature information, and color change feature information; and the geometric feature information comprises edge, profile, vertex, straight line, circle, arc, ellipse, rectangle or curve feature information, or any combination thereof.
- The image processing method according to claim 5, wherein the detecting of the feature information of the obtained images comprises:reading every frame or reading an image captured in real time;detecting feature information of each image; andjudging whether the number of actually detected features is the same as the theoretic number of features; if so, proceeding with subsequent operations; otherwise, discarding a currently detected image, and detecting an image next to the currently detected image.
- The image processing method according to claim 1, wherein: the obtaining of the video images or capturing of a series of the images in real time comprises:obtaining the video images through a video obtaining device while an image with a template is being shaken in front of the video obtaining device; orcapturing a series of images through the video obtaining device in real time while the image with the template is being shaken in front of the video obtaining device.
- The image processing method according to claim 2, wherein: the cropping of the images detected as having the feature information comprises:obtaining coordinates of the feature information of the images according to the obtained feature information;obtaining a maximum value and a minimum value of X-coordinate and Y-coordinate of each point corresponding to the feature information, and obtaining an image suitable for cropping;obtaining a boundary formed by points corresponding to the feature information according to the coordinates of the points corresponding to the feature information of the image;cutting away an area outside the boundary formed by the points and a image suitable for cropping andobtaining a cropped image according to image information inside the boundary formed by the points.
- The image processing method according to claim 1, wherein: the detecting of the difference between the two adjacent images comprises:obtaining information about feature points of the two adjacent images;calculating coordinate difference between the feature points in corresponding positions of the two images;calculating coordinate difference between the feature points in all corresponding positions, and calculating a mean value of them; andjudging whether the mean value of the sum of differences is less than a preset threshold; if so, discarding one of the images because it indicates that the two adjacent images are scarcely different; otherwise, proceeding with subsequent operations.
- The image processing method according to any one of claim 1, claim 2, claim 3, claim 5, claim 6, claim 7, claim 8, or claim 9, wherein: the performing of definition-related treatment for the images detected as having the feature information and comparing of the definition with the preset definition criterion comprises:calculating content of high frequency components of edge information in the feature information of each image detected as having the feature information; andif the content of the high frequency components of edge information in the feature information of the image is greater than the content of the high frequency components corresponding to the preset definition criterion, determining that the image is clear or in focus.
- The image processing method according to claim 4, wherein: the performing of definition-related treatment for the images detected as having the feature information and comparing of the definition with the preset definition criterion comprises:calculating content of high frequency components of edge information in the feature information of each image detected as having the feature information; andif the content of the high frequency components of edge information in the feature information of the image is greater than the content of the high frequency components corresponding to the preset definition criterion, determining that the image is clear or in focus.
- The image processing method according to claim 2 or claim 8, wherein: the performing of definition-related treatment for the images detected as having the feature information, comparing of the definition with the preset definition criterion, and reserving of the clear images comprises:graying the cropped image, calculating a gradient value of the grayed image, and outputting a grayscale image which gives only edge information;setting a corresponding grayscale image threshold as a definition criterion according to the grayscale image;binarizing the cropped image, and obtaining a binarized threshold corresponding to the binarized image; anddetermining that the image is clear and reserving the image if the binarized threshold corresponding to the binarized image is less than the grayscale image threshold.
- The image processing method according to claim 2 or claim 8, wherein: the performing of definition-related treatment for the images detected as having the feature information, comparing of the definition with the preset definition criterion, and reserving of the clear images comprises:graying the cropped image, calculating a gradient value of the grayed image, and outputting a grayscale image which gives only edge information;using a total sum of white pixel points of edge information of the grayscale image or percentage of the edge information to the whole image as a definition criterion;binarizing the cropped image, and obtain the total sum of white pixel points of all edge information in the binarized image or the percentage of the edge information to the whole image; anddetermining that the image is clear and reserving the image if the total sum of the white pixel points of all edge information in the binarized image or the percentage of the edge information to the whole image is less than the definition criterion.
- An image processing system, comprising:an image obtaining module, adapted to obtain video images or capture a series of images in real time;a feature information detecting module, connected with the image obtaining module, and adapted to detect feature information of the obtained video images or the series of images and obtain images that have distinct feature information;an image definition judging module, connected with the feature information detecting module and the image obtaining module, and adapted to: perform definition-related treatment for the images that have the feature information, compare definition of each image with a preset definition criterion, and reserve clear images; andan image difference detecting module, connected with the image definition judging module, and adapted to store a current clear image if difference between the current image and a previous stored image is greater than a preset threshold.
- The image processing system according to claim 14, further comprising:an image cropping module, connected with the feature information detecting module and the image definition judging module, and adapted to crop the images that have the feature information.
- The image processing system according to claim 14, further comprising:an image number judging module, connected with the image difference detecting module and the image obtaining module, and adapted to: detect the number of stored images, and make the image obtaining module select a next image for subsequent operations if the number of stored images does not reach a set number.
- The image processing system according to claim 16, further comprising:a calibrating module, connected with the image number judging module or the image difference detecting module, and adapted to: obtain internal and external parameters of a video obtaining device according to the obtained clear images and the corresponding feature information, and calibrate the video obtaining device.
- An image processing system, comprising:an image obtaining module, adapted to obtain video images or capture a series of images in real time;an image difference detecting module, connected with the image obtaining module, and adapted to store two adjacent images when difference between the two adjacent images is greater than a preset threshold;a feature information detecting module, connected with the image obtaining module, and adapted to detect feature information of the obtained video images or the series of images and obtain images that have distinct feature information; andan image definition judging module, connected with the feature information detecting module and the image obtaining module, and adapted to: perform definition-related treatment for the images that have the feature information, compare definition of each image with a preset definition criterion; when a current image is unclear, discard the current image so that the image obtaining module continues to select a next image.
- The image processing system according to claim 18, further comprising:an image cropping module, connected with the feature information detecting module and the image definition judging module, and adapted to crop the images that have the feature information.
- The image processing system according to claim 18, further comprising:an image number judging module, connected with the image definition judging module and the image obtaining module, and adapted to: detect the number of stored images, and make the image obtaining module select a next image for subsequent operations if the number of stored images does not reach a set number.
- The image processing system according to claim 20, further comprising:a calibrating module, connected with the image number judging module or the image definition judging module, and adapted to: obtain internal and external parameters of a video obtaining device according to the obtained clear images and the corresponding feature information, and calibrate the video obtaining device.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200810082736.2A CN101527040B (en) | 2008-03-05 | 2008-03-05 | Method and system for processing images |
PCT/CN2009/070583 WO2009109125A1 (en) | 2008-03-05 | 2009-02-27 | Image processing method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
EP2252088A1 true EP2252088A1 (en) | 2010-11-17 |
EP2252088A4 EP2252088A4 (en) | 2011-04-27 |
Family
ID=41055550
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP09718449A Withdrawn EP2252088A4 (en) | 2008-03-05 | 2009-02-27 | Image processing method and system |
Country Status (4)
Country | Link |
---|---|
US (1) | US8416314B2 (en) |
EP (1) | EP2252088A4 (en) |
CN (1) | CN101527040B (en) |
WO (1) | WO2009109125A1 (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102098411A (en) * | 2010-12-20 | 2011-06-15 | 东莞市金翔电器设备有限公司 | High-velocity scanning method capable of realizing compound sensing |
CN103927750A (en) * | 2014-04-18 | 2014-07-16 | 上海理工大学 | Detection method of checkboard grid image angular point sub pixel |
Families Citing this family (39)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101527040B (en) | 2008-03-05 | 2012-12-19 | 华为终端有限公司 | Method and system for processing images |
CN102098379A (en) * | 2010-12-17 | 2011-06-15 | 惠州Tcl移动通信有限公司 | Terminal as well as method and device for acquiring real-time video images of terminal |
KR101366860B1 (en) * | 2011-09-20 | 2014-02-21 | 엘지전자 주식회사 | Mobile robot and controlling method of the same |
CN103093451A (en) * | 2011-11-03 | 2013-05-08 | 北京理工大学 | Checkerboard intersection recognition algorithm |
CN104427333A (en) * | 2013-08-20 | 2015-03-18 | 北京市博汇科技股份有限公司 | High-definition television signal detecting method and system |
US9646224B2 (en) | 2014-04-14 | 2017-05-09 | Shenzhen China Star Optoelectronics Technology Co., Ltd. | Image processing method, image processing device and automated optical inspection machine |
CN103927749A (en) * | 2014-04-14 | 2014-07-16 | 深圳市华星光电技术有限公司 | Image processing method and device and automatic optical detector |
CN104936002B (en) * | 2015-06-05 | 2018-11-06 | 网易有道信息技术(北京)有限公司 | A kind of method and apparatus of screen recording |
CN107026827B (en) * | 2016-02-02 | 2020-04-03 | 上海交通大学 | Optimized transmission method for static image in video stream |
CN106231201B (en) * | 2016-08-31 | 2020-06-05 | 成都极米科技股份有限公司 | Automatic focusing method and device |
CN106651908B (en) * | 2016-10-13 | 2020-03-31 | 北京科技大学 | Multi-moving-target tracking method |
CN106454094A (en) * | 2016-10-19 | 2017-02-22 | 广东欧珀移动通信有限公司 | Shooting method and device, and mobile terminal |
CN108713318A (en) * | 2016-10-31 | 2018-10-26 | 华为技术有限公司 | A kind of processing method and equipment of video frame |
CN106789555A (en) * | 2016-11-25 | 2017-05-31 | 努比亚技术有限公司 | Method of transmitting video data and device |
CN106990518A (en) * | 2017-04-17 | 2017-07-28 | 深圳大学 | A kind of blood film self-focusing micro imaging method |
CN111034199A (en) * | 2017-06-30 | 2020-04-17 | 诺基亚通信公司 | Real-time video |
CN107358221B (en) * | 2017-08-08 | 2020-10-09 | 大连万和海拓文化体育产业有限公司 | Video recognition technology-based chessboard positioning method for automatically recording chess manual of go |
CN107506701B (en) * | 2017-08-08 | 2021-03-05 | 大连万和海拓文化体育产业有限公司 | Automatic go chess manual recording method based on video recognition technology |
CN107610131B (en) * | 2017-08-25 | 2020-05-12 | 百度在线网络技术(北京)有限公司 | Image clipping method and image clipping device |
CN107610108B (en) * | 2017-09-04 | 2019-04-26 | 腾讯科技(深圳)有限公司 | Image processing method and device |
CN109800619B (en) * | 2017-11-16 | 2022-07-01 | 湖南生物机电职业技术学院 | Image recognition method for citrus fruits in mature period |
JP7353747B2 (en) * | 2018-01-12 | 2023-10-02 | キヤノン株式会社 | Information processing device, system, method, and program |
CN108596874B (en) * | 2018-03-17 | 2024-01-05 | 紫光汇智信息技术有限公司 | Image definition judging method and device, computer equipment and product |
CN110971811B (en) * | 2018-09-30 | 2022-11-15 | 中兴通讯股份有限公司 | Image screening method, system, terminal and computer readable storage medium |
CN111091526B (en) * | 2018-10-23 | 2023-06-13 | 广州弘度信息科技有限公司 | Video blurring detection method and system |
CN109598764B (en) * | 2018-11-30 | 2021-07-09 | Oppo广东移动通信有限公司 | Camera calibration method and device, electronic equipment and computer-readable storage medium |
CN109493391A (en) * | 2018-11-30 | 2019-03-19 | Oppo广东移动通信有限公司 | Camera calibration method and device, electronic equipment, computer readable storage medium |
CN109862390B (en) * | 2019-02-26 | 2021-06-01 | 北京融链科技有限公司 | Method and device for optimizing media stream, storage medium and processor |
CN110049247B (en) * | 2019-04-28 | 2021-03-12 | Oppo广东移动通信有限公司 | Image optimization method and device, electronic equipment and readable storage medium |
CN110298229B (en) * | 2019-04-29 | 2022-04-01 | 星河视效科技(北京)有限公司 | Video image processing method and device |
CN110517246B (en) * | 2019-08-23 | 2022-04-08 | 腾讯科技(深圳)有限公司 | Image processing method and device, electronic equipment and storage medium |
CN110782647A (en) * | 2019-11-06 | 2020-02-11 | 重庆神缘智能科技有限公司 | Intelligent meter reading system based on image recognition |
CN111161332A (en) * | 2019-12-30 | 2020-05-15 | 上海研境医疗科技有限公司 | Homologous pathology image registration preprocessing method, device, equipment and storage medium |
CN113330487A (en) * | 2019-12-30 | 2021-08-31 | 华为技术有限公司 | Parameter calibration method and device |
US11043007B1 (en) * | 2020-05-26 | 2021-06-22 | Black Sesame International Holding Limited | Dual camera calibration |
CN111639708B (en) * | 2020-05-29 | 2023-05-09 | 深圳市燕麦科技股份有限公司 | Image processing method, device, storage medium and equipment |
CN112351196B (en) * | 2020-09-22 | 2022-03-11 | 北京迈格威科技有限公司 | Image definition determining method, image focusing method and device |
CN112989113B (en) * | 2021-04-21 | 2021-08-31 | 北京金和网络股份有限公司 | Picture screening method, device and equipment |
CN113286079B (en) * | 2021-05-10 | 2023-04-28 | 迈克医疗电子有限公司 | Image focusing method and device, electronic equipment and readable storage medium |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
GB2195214A (en) * | 1980-12-10 | 1988-03-30 | Emi Ltd | Automatically focussing system for an optical system |
US6768509B1 (en) * | 2000-06-12 | 2004-07-27 | Intel Corporation | Method and apparatus for determining points of interest on an image of a camera calibration object |
Family Cites Families (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6956779B2 (en) * | 1999-01-14 | 2005-10-18 | Silicon Storage Technology, Inc. | Multistage autozero sensing for a multilevel non-volatile memory integrated circuit system |
GB2370438A (en) * | 2000-12-22 | 2002-06-26 | Hewlett Packard Co | Automated image cropping using selected compositional rules. |
KR20040065928A (en) * | 2003-01-16 | 2004-07-23 | 삼성전자주식회사 | Printer having printing function of moving picture and control method thereof |
CN1567370A (en) * | 2003-06-23 | 2005-01-19 | 威视科技股份有限公司 | video processing method and device |
US8896725B2 (en) * | 2007-06-21 | 2014-11-25 | Fotonation Limited | Image capture device with contemporaneous reference image capture mechanism |
US7454040B2 (en) * | 2003-08-29 | 2008-11-18 | Hewlett-Packard Development Company, L.P. | Systems and methods of detecting and correcting redeye in an image suitable for embedded applications |
JP4318553B2 (en) * | 2004-01-23 | 2009-08-26 | 三洋電機株式会社 | Image signal processing device |
JP4324038B2 (en) * | 2004-07-07 | 2009-09-02 | Okiセミコンダクタ株式会社 | YC separation circuit |
US20060023077A1 (en) * | 2004-07-30 | 2006-02-02 | Microsoft Corporation | System and method for photo editing |
US20060114327A1 (en) * | 2004-11-26 | 2006-06-01 | Fuji Photo Film, Co., Ltd. | Photo movie creating apparatus and program |
JP4241709B2 (en) * | 2005-10-11 | 2009-03-18 | ソニー株式会社 | Image processing device |
CN101419705B (en) | 2007-10-24 | 2011-01-05 | 华为终端有限公司 | Video camera demarcating method and device |
CN101527040B (en) | 2008-03-05 | 2012-12-19 | 华为终端有限公司 | Method and system for processing images |
US8300929B2 (en) * | 2009-10-07 | 2012-10-30 | Seiko Epson Corporation | Automatic red-eye object classification in digital photographic images |
-
2008
- 2008-03-05 CN CN200810082736.2A patent/CN101527040B/en active Active
-
2009
- 2009-02-27 WO PCT/CN2009/070583 patent/WO2009109125A1/en active Application Filing
- 2009-02-27 EP EP09718449A patent/EP2252088A4/en not_active Withdrawn
-
2010
- 2010-08-20 US US12/860,339 patent/US8416314B2/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
GB2195214A (en) * | 1980-12-10 | 1988-03-30 | Emi Ltd | Automatically focussing system for an optical system |
US6768509B1 (en) * | 2000-06-12 | 2004-07-27 | Intel Corporation | Method and apparatus for determining points of interest on an image of a camera calibration object |
Non-Patent Citations (3)
Title |
---|
ALBERT C. LIM ET AL.: "Image resolution and performance analysis of webcams for ground-based astronomy", PROC. SPIE GROUND-BASED TELESCOPES, vol. 5489, no. 1152, 1163, 21 June 2004 (2004-06-21), XP040188036, * |
MCBRIDE J ET AL: "Single Camera Stereo for Mobile Robot Surveillance", COMPUTER VISION AND PATTERN RECOGNITION, 2005 IEEE COMPUTER SOCIETY CONFERENCE ON, IEEE, PISCATAWAY, NJ, USA, 25 June 2005 (2005-06-25), page 128, XP031259772, ISBN: 978-0-7695-2372-9 * |
See also references of WO2009109125A1 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102098411A (en) * | 2010-12-20 | 2011-06-15 | 东莞市金翔电器设备有限公司 | High-velocity scanning method capable of realizing compound sensing |
CN102098411B (en) * | 2010-12-20 | 2012-11-14 | 东莞市金翔电器设备有限公司 | High-velocity scanning method capable of realizing compound sensing |
CN103927750A (en) * | 2014-04-18 | 2014-07-16 | 上海理工大学 | Detection method of checkboard grid image angular point sub pixel |
CN103927750B (en) * | 2014-04-18 | 2016-09-14 | 上海理工大学 | The detection method of gridiron pattern image angular-point sub-pixel |
Also Published As
Publication number | Publication date |
---|---|
CN101527040A (en) | 2009-09-09 |
US8416314B2 (en) | 2013-04-09 |
WO2009109125A1 (en) | 2009-09-11 |
CN101527040B (en) | 2012-12-19 |
EP2252088A4 (en) | 2011-04-27 |
US20100315512A1 (en) | 2010-12-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8416314B2 (en) | Method and system for processing images | |
CN111028189B (en) | Image processing method, device, storage medium and electronic equipment | |
EP2849431B1 (en) | Method and apparatus for detecting backlight | |
JP5592006B2 (en) | 3D image processing | |
WO2019233264A1 (en) | Image processing method, computer readable storage medium, and electronic device | |
CN113365041A (en) | Projection correction method, projection correction device, storage medium and electronic equipment | |
WO2006018056A1 (en) | Method and apparatus for detecting and correcting red-eye effect | |
KR101692227B1 (en) | A panorama image generation method using FAST algorithm | |
CN107493407B (en) | Photographing device and photographing method | |
CN109738450B (en) | Method and device for detecting notebook keyboard | |
EP3627822B1 (en) | Focus region display method and apparatus, and terminal device | |
CN111368819B (en) | Light spot detection method and device | |
CN109559353B (en) | Camera module calibration method and device, electronic equipment and computer readable storage medium | |
CN111695540A (en) | Video frame identification method, video frame cutting device, electronic equipment and medium | |
CN112272292A (en) | Projection correction method, apparatus and storage medium | |
CN112261292B (en) | Image acquisition method, terminal, chip and storage medium | |
CN113822927B (en) | Face detection method, device, medium and equipment suitable for weak quality image | |
US20170263013A1 (en) | Information processing apparatus, information processing method, and storage medium | |
CN111885371A (en) | Image occlusion detection method and device, electronic equipment and computer readable medium | |
US20210281742A1 (en) | Document detections from video images | |
CN108647680B (en) | Image positioning frame detection method and device | |
WO2022127491A1 (en) | Image processing method and device, and storage medium and terminal | |
CN113516595B (en) | Image processing method, image processing apparatus, electronic device, and storage medium | |
US7376285B2 (en) | Method of auto-deskewing a tilted image | |
CN111630569B (en) | Binocular matching method, visual imaging device and device with storage function |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
17P | Request for examination filed |
Effective date: 20100826 |
|
AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO SE SI SK TR |
|
AX | Request for extension of the european patent |
Extension state: AL BA RS |
|
A4 | Supplementary search report drawn up and despatched |
Effective date: 20110324 |
|
DAX | Request for extension of the european patent (deleted) | ||
17Q | First examination report despatched |
Effective date: 20120216 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN |
|
18D | Application deemed to be withdrawn |
Effective date: 20131213 |